cuda-cccl 0.3.2__cp313-cp313-manylinux_2_24_aarch64.whl → 0.3.4__cp313-cp313-manylinux_2_24_aarch64.whl

cuda/cccl/headers/include/thrust/system/cpp/detail/memory.inl

@@ -31,11 +31,8 @@
 #include <cuda/std/limits>
 
 THRUST_NAMESPACE_BEGIN
-namespace system
+namespace system::cpp
 {
-namespace cpp
-{
-
 pointer<void> malloc(std::size_t n)
 {
   tag t;
@@ -54,7 +51,5 @@ void free(pointer<void> ptr)
   tag t;
   return thrust::system::detail::sequential::free(t, ptr);
 } // end free()
-
-} // namespace cpp
-} // namespace system
+} // namespace system::cpp
 THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/cpp/memory.h

@@ -39,7 +39,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace system::cpp
 {
-
 /*! Allocates an area of memory available to Thrust's <tt>cpp</tt> system.
  *  \param n Number of bytes to allocate.
  *  \return A <tt>cpp::pointer<void></tt> pointing to the beginning of the newly

cuda/cccl/headers/include/thrust/system/cpp/memory_resource.h

@@ -36,7 +36,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace system::cpp
 {
-
 //! \cond
 namespace detail
 {
@@ -66,7 +65,6 @@ using universal_host_pinned_memory_resource = universal_memory_resource;
 
 /*! \} // memory_resources
  */
-
 } // namespace system::cpp
 
 THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/cpp/pointer.h

@@ -37,7 +37,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace system::cpp
 {
-
 /*! \p cpp::pointer stores a pointer to an object allocated in memory accessible
  *  by the \p cpp system. This type provides type safety when dispatching
  *  algorithms on ranges resident in \p cpp memory.
@@ -99,7 +98,6 @@ using universal_host_pinned_pointer = universal_pointer<T>;
  */
 template <typename T>
 using reference = thrust::reference<T, thrust::system::cpp::tag>;
-
 } // namespace system::cpp
 
 /*! \addtogroup system_backends Systems

cuda/cccl/headers/include/thrust/system/cpp/vector.h

@@ -38,7 +38,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace system::cpp
 {
-
 /*! \p cpp::vector is a container that supports random access to elements,
  *  constant time removal of elements at the end, and linear time insertion
  *  and removal of elements at the beginning or in the middle. The number of

cuda/cccl/headers/include/thrust/system/cuda/detail/adjacent_difference.h

@@ -68,10 +68,8 @@ _CCCL_HOST_DEVICE OutputIterator adjacent_difference(
 
 namespace cuda_cub
 {
-
 namespace __adjacent_difference
 {
-
 template <cub::MayAlias AliasOpt, class InputIt, class OutputIt, class BinaryOp>
 cudaError_t THRUST_RUNTIME_FUNCTION doit_step(
   void* d_temp_storage,
@@ -184,7 +182,6 @@ adjacent_difference(execution_policy<Derived>& policy, InputIt first, InputIt la
 
   return result + num_items;
 }
-
 } // namespace __adjacent_difference
 
 //-------------------------
@@ -209,7 +206,6 @@ adjacent_difference(execution_policy<Derived>& policy, InputIt first, InputIt la
   using input_type = thrust::detail::it_value_t<InputIt>;
   return cuda_cub::adjacent_difference(policy, first, last, result, ::cuda::std::minus<input_type>());
 }
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END
 

cuda/cccl/headers/include/thrust/system/cuda/detail/copy.h

@@ -268,6 +268,5 @@ copy_n(cross_system<System1, System2> systems, InputIterator first, Size n, Outp
 {
   return __copy::cross_system_copy_n(systems, first, n, result);
 }
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/cuda/detail/copy_if.h

@@ -84,10 +84,8 @@ _CCCL_HOST_DEVICE OutputIterator copy_if(
 
 namespace cuda_cub
 {
-
 namespace detail
 {
-
 template <cub::SelectImpl SelectionOpt,
           typename Derived,
           typename InputIt,
@@ -218,7 +216,6 @@ THRUST_RUNTIME_FUNCTION OutputIt copy_if(
 
   return output;
 }
-
 } // namespace detail
 
 //-------------------------
@@ -247,7 +244,6 @@ OutputIterator _CCCL_HOST_DEVICE copy_if(
   THRUST_CDP_DISPATCH((return detail::copy_if<cub::SelectImpl::Select>(policy, first, last, stencil, result, pred);),
                       (return thrust::copy_if(cvt_to_seq(derived_cast(policy)), first, last, stencil, result, pred);));
 }
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END
 

cuda/cccl/headers/include/thrust/system/cuda/detail/core/agent_launcher.h

@@ -58,11 +58,7 @@ _CCCL_DIAG_SUPPRESS_NVHPC(attribute_requires_external_linkage)
 
 THRUST_NAMESPACE_BEGIN
 
-namespace cuda_cub
-{
-namespace core
-{
-namespace detail
+namespace cuda_cub::core::detail
 {
 #  ifndef THRUST_DETAIL_KERNEL_ATTRIBUTES
 #    define THRUST_DETAIL_KERNEL_ATTRIBUTES CCCL_DETAIL_KERNEL_ATTRIBUTES
@@ -279,10 +275,7 @@ struct AgentLauncher : Agent
     sync();
   }
 };
-
-} // namespace detail
-} // namespace core
-} // namespace cuda_cub
+} // namespace cuda_cub::core::detail
 
 THRUST_NAMESPACE_END
 

cuda/cccl/headers/include/thrust/system/cuda/detail/core/triple_chevron_launch.h

@@ -1,29 +1,5 @@
-/******************************************************************************
- * Copyright (c) 2024, NVIDIA CORPORATION.  All rights reserved.
- *
- * Redistribution and use in source and binary forms, with or without
- * modification, are permitted provided that the following conditions are met:
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in the
- *       documentation and/or other materials provided with the distribution.
- *     * Neither the name of the NVIDIA CORPORATION nor the
- *       names of its contributors may be used to endorse or promote products
- *       derived from this software without specific prior written permission.
- *
- * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
- * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
- * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
- * ARE DISCLAIMED. IN NO EVENT SHALL NVIDIA CORPORATION BE LIABLE FOR ANY
- * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
- * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
- * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
- * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
- * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
- *
- ******************************************************************************/
+// SPDX-FileCopyrightText: Copyright (c) 2024, NVIDIA CORPORATION. All rights reserved.
+// SPDX-License-Identifier: BSD-3
 #pragma once
 
 #include <thrust/detail/config.h>
@@ -43,11 +19,8 @@
 
 THRUST_NAMESPACE_BEGIN
 
-namespace cuda_cub
+namespace cuda_cub::detail
 {
-namespace detail
-{
-
 struct _CCCL_VISIBILITY_HIDDEN triple_chevron
 {
   using Size = size_t;
@@ -185,7 +158,6 @@ struct _CCCL_VISIBILITY_HIDDEN triple_chevron
   }
 
 }; // struct triple_chevron
-} // namespace detail
-} // namespace cuda_cub
+} // namespace cuda_cub::detail
 
 THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/cuda/detail/core/util.h

@@ -45,11 +45,7 @@
 
 THRUST_NAMESPACE_BEGIN
 
-namespace cuda_cub
-{
-namespace core
-{
-namespace detail
+namespace cuda_cub::core::detail
 {
 /// Typelist - a container of types
 template <typename...>
@@ -585,9 +581,6 @@ THRUST_RUNTIME_FUNCTION cudaError_t alias_storage(
   return cub::detail::AliasTemporaries(storage_ptr, storage_size, allocations, allocation_sizes);
 }
 #endif // !_CCCL_COMPILER(NVRTC)
-
-} // namespace detail
-} // namespace core
-} // namespace cuda_cub
+} // namespace cuda_cub::core::detail
 
 THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/cuda/detail/count.h

@@ -47,7 +47,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace cuda_cub
 {
-
 template <class Derived, class InputIt, class UnaryPred>
 thrust::detail::it_difference_t<InputIt> _CCCL_HOST_DEVICE
 count_if(execution_policy<Derived>& policy, InputIt first, InputIt last, UnaryPred unary_pred)
@@ -69,7 +68,6 @@ count(execution_policy<Derived>& policy, InputIt first, InputIt last, Value cons
 {
   return cuda_cub::count_if(policy, first, last, thrust::detail::equal_to_value<Value>{value});
 }
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END
 #endif

cuda/cccl/headers/include/thrust/system/cuda/detail/cross_system.h

@@ -43,7 +43,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace cuda_cub
 {
-
 template <class Sys1, class Sys2>
 struct cross_system : execution_policy<cross_system<Sys1, Sys2>>
 {
@@ -238,6 +237,5 @@ select_system(thrust::cpp::execution_policy<Sys1> const& sys1, execution_policy<
   thrust::execution_policy<Sys2>& non_const_sys2      = const_cast<execution_policy<Sys2>&>(sys2);
   return cross_system<Sys1, Sys2>(non_const_sys1, non_const_sys2);
 }
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/cuda/detail/dispatch.h

@@ -26,16 +26,37 @@
 #  pragma system_header
 #endif // no system header
 
-#include <thrust/detail/integer_math.h>
 #include <thrust/detail/preprocessor.h>
 
+#include <cuda/std/__concepts/concept_macros.h>
+#include <cuda/std/__exception/throw_error.h>
+#include <cuda/std/__type_traits/is_arithmetic.h>
+#include <cuda/std/__type_traits/is_unsigned.h>
 #include <cuda/std/cstdint>
-#include <cuda/std/detail/libcxx/include/stdexcept>
 #include <cuda/std/limits>
 #include <cuda/std/type_traits>
 
 #include <string>
 
+THRUST_NAMESPACE_BEGIN
+namespace detail
+{
+_CCCL_TEMPLATE(typename T)
+_CCCL_REQUIRES(::cuda::std::is_arithmetic_v<T>)
+[[nodiscard]] _CCCL_API constexpr bool is_negative([[maybe_unused]] T x) noexcept
+{
+  if constexpr (::cuda::std::is_unsigned_v<T>)
+  {
+    return false;
+  }
+  else
+  {
+    return x < 0;
+  }
+}
+} // namespace detail
+THRUST_NAMESPACE_END
+
 #if defined(THRUST_FORCE_32_BIT_OFFSET_TYPE) && defined(THRUST_FORCE_64_BIT_OFFSET_TYPE)
 #  error "Only THRUST_FORCE_32_BIT_OFFSET_TYPE or THRUST_FORCE_64_BIT_OFFSET_TYPE may be defined!"
 #endif // THRUST_FORCE_32_BIT_OFFSET_TYPE && THRUST_FORCE_64_BIT_OFFSET_TYPE

cuda/cccl/headers/include/thrust/system/cuda/detail/equal.h

@@ -44,7 +44,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace cuda_cub
 {
-
 template <class Derived, class InputIt1, class InputIt2, class BinaryPred>
 bool _CCCL_HOST_DEVICE
 equal(execution_policy<Derived>& policy, InputIt1 first1, InputIt1 last1, InputIt2 first2, BinaryPred binary_pred)
@@ -58,7 +57,6 @@ bool _CCCL_HOST_DEVICE equal(execution_policy<Derived>& policy, InputIt1 first1,
   using InputType1 = thrust::detail::it_value_t<InputIt1>;
   return cuda_cub::equal(policy, first1, last1, first2, ::cuda::std::equal_to<InputType1>());
 }
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END
 #endif

cuda/cccl/headers/include/thrust/system/cuda/detail/error.inl

@@ -32,7 +32,6 @@ THRUST_NAMESPACE_BEGIN
 
 namespace system
 {
-
 error_code make_error_code(cuda::errc::errc_t e)
 {
   return error_code(static_cast<int>(e), cuda_category());
@@ -43,12 +42,8 @@ error_condition make_error_condition(cuda::errc::errc_t e)
   return error_condition(static_cast<int>(e), cuda_category());
 } // end make_error_condition()
 
-namespace cuda_cub
-{
-
-namespace detail
+namespace cuda_cub::detail
 {
-
 class cuda_error_category : public error_category
 {
 public:
@@ -80,17 +75,13 @@ public:
     return system_category().default_error_condition(ev);
   }
 }; // end cuda_error_category
-
-} // namespace detail
-
-} // end namespace cuda_cub
+} // end namespace cuda_cub::detail
 
 const error_category& cuda_category()
 {
   static const thrust::system::cuda_cub::detail::cuda_error_category result;
   return result;
 }
-
 } // end namespace system
 
 THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/cuda/detail/execution_policy.h

@@ -26,6 +26,8 @@ namespace cuda_cub
 //! \addtogroup execution_policies
 //! \{
 
+// note: the tag and execution policy need to be defined in the same namespace as the algorithms for ADL to find them
+
 //! \p thrust::cuda::tag is a type representing Thrust's CUDA backend system in C++'s type system.
 //! Iterators "tagged" with a type which is convertible to \p cuda::tag assert that they may be
 //! "dispatched" to algorithm implementations in the \p cuda system.

cuda/cccl/headers/include/thrust/system/cuda/detail/extrema.h

@@ -57,10 +57,8 @@
 THRUST_NAMESPACE_BEGIN
 namespace cuda_cub
 {
-
 namespace __extrema
 {
-
 template <class InputType, class IndexType, class Predicate>
 struct arg_min_f
 {
@@ -386,7 +384,6 @@ element(execution_policy<Derived>& policy, ItemsIt first, ItemsIt last, BinaryPr
   T result = extrema(policy, begin, num_items, arg_min_t(binary_pred), (T*) (nullptr));
   return first + thrust::get<1>(result);
 }
-
 } // namespace __extrema
 
 /// min element
@@ -470,7 +467,6 @@ pair<ItemsIt, ItemsIt> _CCCL_HOST_DEVICE minmax_element(execution_policy<Derived
   using value_type = thrust::detail::it_value_t<ItemsIt>;
   return cuda_cub::minmax_element(policy, first, last, ::cuda::std::less<value_type>());
 }
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END
 #endif

cuda/cccl/headers/include/thrust/system/cuda/detail/fill.h

@@ -94,7 +94,6 @@ fill(execution_policy<Derived>& policy, ForwardIterator first, ForwardIterator l
 {
   cuda_cub::fill_n(policy, first, ::cuda::std::distance(first, last), value);
 }
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END
 #endif

cuda/cccl/headers/include/thrust/system/cuda/detail/find.h

@@ -49,7 +49,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace cuda_cub
 {
-
 // XXX forward declare to circumvent circular dependency
 template <class Derived, class InputIt, class Predicate>
 InputIt _CCCL_HOST_DEVICE find_if(execution_policy<Derived>& policy, InputIt first, InputIt last, Predicate predicate);
@@ -60,7 +59,6 @@ find_if_not(execution_policy<Derived>& policy, InputIt first, InputIt last, Pred
 
 template <class Derived, class InputIt, class T>
 InputIt _CCCL_HOST_DEVICE find(execution_policy<Derived>& policy, InputIt first, InputIt last, T const& value);
-
 }; // namespace cuda_cub
 THRUST_NAMESPACE_END
 
@@ -69,10 +67,8 @@ THRUST_NAMESPACE_END
 THRUST_NAMESPACE_BEGIN
 namespace cuda_cub
 {
-
 namespace __find_if
 {
-
 template <typename TupleType>
 struct functor
 {
@@ -164,7 +160,6 @@ InputIt _CCCL_HOST_DEVICE find(execution_policy<Derived>& policy, InputIt first,
 
   return cuda_cub::find_if(policy, first, last, _1 == value);
 }
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END
 #endif

cuda/cccl/headers/include/thrust/system/cuda/detail/for_each.h

@@ -51,7 +51,6 @@ THRUST_NAMESPACE_BEGIN
 
 namespace cuda_cub
 {
-
 // for_each_n
 _CCCL_EXEC_CHECK_DISABLE
 template <class Derived, class Input, class Size, class UnaryOp>

cuda/cccl/headers/include/thrust/system/cuda/detail/gather.h

@@ -45,7 +45,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace cuda_cub
 {
-
 template <class Derived, class MapIt, class ItemsIt, class ResultIt>
 ResultIt _CCCL_HOST_DEVICE
 gather(execution_policy<Derived>& policy, MapIt map_first, MapIt map_last, ItemsIt items, ResultIt result)
@@ -84,7 +83,6 @@ ResultIt _CCCL_HOST_DEVICE gather_if(
 {
   return cuda_cub::gather_if(policy, map_first, map_last, stencil, items, result, ::cuda::std::identity{});
 }
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END
 

cuda/cccl/headers/include/thrust/system/cuda/detail/generate.h

@@ -23,7 +23,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace cuda_cub
 {
-
 _CCCL_EXEC_CHECK_DISABLE
 template <class Derived, class OutputIt, class Size, class Generator>
 OutputIt _CCCL_HOST_DEVICE
@@ -54,7 +53,6 @@ void _CCCL_HOST_DEVICE generate(execution_policy<Derived>& policy, OutputIt firs
 {
   cuda_cub::generate_n(policy, first, ::cuda::std::distance(first, last), generator);
 }
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END
 #endif

cuda/cccl/headers/include/thrust/system/cuda/detail/iter_swap.h

@@ -74,7 +74,6 @@ inline _CCCL_HOST_DEVICE void iter_swap(thrust::cuda::execution_policy<DerivedPo
                 swap(*thrust::raw_pointer_cast(a), *thrust::raw_pointer_cast(b));));
 
 } // end iter_swap()
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END
 #endif

cuda/cccl/headers/include/thrust/system/cuda/detail/make_unsigned_special.h

@@ -27,12 +27,8 @@
 #endif // no system header
 
 THRUST_NAMESPACE_BEGIN
-namespace cuda_cub
+namespace cuda_cub::detail
 {
-
-namespace detail
-{
-
 template <typename Size>
 struct make_unsigned_special;
 
@@ -55,7 +51,5 @@ struct make_unsigned_special<long long>
 {
   using type = unsigned long long;
 };
-
-} // namespace detail
-} // namespace cuda_cub
+} // namespace cuda_cub::detail
 THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/cuda/detail/malloc_and_free.h

@@ -45,7 +45,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace cuda_cub
 {
-
 #ifdef THRUST_CACHING_DEVICE_MALLOC
 #  define __CUB_CACHING_MALLOC
 #  if _CCCL_HOST_COMPILATION()
@@ -116,6 +115,5 @@ _CCCL_HOST_DEVICE void free(execution_policy<DerivedPolicy>&, Pointer ptr)
                  thrust::free(thrust::seq, ptr);));
 #endif
 } // end free()
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/cuda/detail/memory.inl

@@ -33,7 +33,6 @@
 THRUST_NAMESPACE_BEGIN
 namespace cuda_cub
 {
-
 _CCCL_HOST_DEVICE pointer<void> malloc(std::size_t n)
 {
   tag cuda_tag;
@@ -52,6 +51,5 @@ _CCCL_HOST_DEVICE void free(pointer<void> ptr)
   tag cuda_tag;
   return thrust::cuda_cub::free(cuda_tag, ptr.get());
 } // end free()
-
 } // namespace cuda_cub
 THRUST_NAMESPACE_END

cuda/cccl/headers/include/thrust/system/cuda/detail/merge.h

@@ -1,29 +1,5 @@
-/******************************************************************************
- * Copyright (c) 2024, NVIDIA CORPORATION.  All rights reserved.
- *
- * Redistribution and use in source and binary forms, with or without
- * modification, are permitted provided that the following conditions are met:
- *     * Redistributions of source code must retain the above copyright
- *       notice, this list of conditions and the following disclaimer.
- *     * Redistributions in binary form must reproduce the above copyright
- *       notice, this list of conditions and the following disclaimer in the
- *       documentation and/or other materials provided with the distribution.
- *     * Neither the name of the NVIDIA CORPORATION nor the
- *       names of its contributors may be used to endorse or promote products
- *       derived from this software without specific prior written permission.
- *
- * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
- * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
- * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
- * ARE DISCLAIMED. IN NO EVENT SHALL NVIDIA CORPORATION BE LIABLE FOR ANY
- * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
- * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
- * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
- * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
- * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
- *
- ******************************************************************************/
+// SPDX-FileCopyrightText: Copyright (c) 2024, NVIDIA CORPORATION. All rights reserved.
+// SPDX-License-Identifier: BSD-3
 #pragma once
 
 #include <thrust/detail/config.h>